An extensive background on electrostatic discharge (ESD) may be found on the Internet at the website of the ESD Association, 7900 Turin Road, Building 3, Suite 2, Rome, N.Y. 13440-2069. See xe2x80x9cwww.esda.orgxe2x80x9d for this information.
A number of ESD dissipating tools and containers have been devised and used to prevent ESD events and protect sensitive microelectronic devices. It is known that the ESD dissipating materials should not be fully insulative or conductive. Rather such materials should be semi-insulative (having a Volume Resistivity or xe2x80x9cRvxe2x80x9d ranging from 103-1011 Ohm-cm). For example, ESD dissipating materials have been suggested in manufacturing and testing MR/GMR heads and assembling hard disk drives.1,2 Similarly, ESD dissipating ceramic and carbon loaded polymer tweezers have been used in handling semiconductor parts.
The use of semi-insulative ceramics has been reported for some ESD control applications. In general, semi-insulative ceramics were prepared by mixing conductive or semiconductive ceramics with insulative ceramics, which is a similar manufacturing method to those of ESD dissipative polymers. In the past, some semi-insulative materials were prepared for other applications based on the mixing rule. For example, U.S. Pat. No. 3,410,728 reports that metal oxide doped stabilized zirconia have both ionic and electric conductivities at elevated temperature. LaCrO3 doped stabilized zirconia3 has also been tried for a high temperature electrical conductor.
Recently, German Patent No. 3743630 C1 (1989)4 reported that tweezers made out of various ceramics having surface resistances of 105-1010 Ohms can be used to hold electrostatically sensitive components to protect against electrostatic discharge. The patent further indicated that partially or fully stabilized, zirconia based ceramics can be used as a base material. However, the patent did not disclose any information about manufacturing process and other physical properties.
Japanese Patent Application No. JP62-25404 A (1987)5 reported that the resistivity of stabilized zirconia can be controlled in the range of 7xc3x97102-5.5xc3x97105 Ohm (the resistance value can not be converted to resistivity without electrode dimensions) at 600xc2x0 C. by adding 5-50 mol. % of transition metal oxide (Co2O3) for thermistor elements. Again, the patent did not disclose any information about other physical properties.
Japanese Patent Application No. JP3-5363 A (1991) reported that ceramic tape guides made from a mixture of TiO2 (50-99 wt. %) and Al2O3 can be heat treated in various atmosphere to volume resistivities of 104-1011 Ohm-cm to protect the magnetic tape drive from electrostatic discharge damage. The tape guide is more resistant against wear due to its higher hardness of 700-900 Kg/mm2 compared to conventional tape guides made out of stainless steel.
U.S. Pat. No. 5,612,144 (1997)6 reports that alumina and silicon carbide based ceramics having resistivities of 2xc3x97106 to 1010 Ohm/cm2 can be prepared by adding nitrides and carbides of Ti, Zr, Hf, Nb and Ta for electrification removing components.
U.S. Pat. No. 5,830,819 (1998)7 reported alumina composite ceramics with volume resistivities from 1xc3x97107 to 1xc3x971013 Ohm-cm at 25-75xc2x0 C. and an absolute value of the temperature coefficient of volume resistivity (TCR) of not larger than 1.8%/xc2x0 C. can be prepared with additives containing transition metal oxides such as TiO2, Cr2O3, Co3O4, MnO2, NiO, Fe2O3, and V2O5 for antistatic part such as conveyer arm, handling jig, tweezers for holding wafers.
U.S. Pat. No. 5,958,813 (1999)8 reports that semi-insulating aluminum nitride (AlN) sintered body can be made to the resistivity of 104 to 1011 by forming an intergranular phase from at least one oxides of Ti, Ce, Ni, Ta, Y, Er and Yb, or from Si to prepare a member for removing static electricity.
Most materials described above are not necessarily structurally reliable for high performance ESD dissipative ceramic tools because of low flexural strength, fracture toughness and residual porosities in their microstructure. PCT Patent Publication No. WO 98/49121 (1998) reports that semiconductive zirconia with good mechanical properties can be prepared using 10 to 40 wt. % oxides of Fe, Co, Ni, and Cr.
Japanese Patent Application No. JP8-39441 (1996) reported that tweezers with volume resistivities of 5xc3x97107 to 1xc3x97109 Ohm-cm can be used to hold semiconductor parts to avoid ESD problems.
C. Lam (1996)9 has reported that the ESD dissipative polymers and ceramic tweezers can be used during the manufacturing and handling of MR heads. The test result indicated that tweezers made out of a xe2x80x9cdopedxe2x80x9d zirconia (vendor proprietary) showed the best performance.
Japanese Patent Application No. JP 10-296646 A (1998) reported a high strength, zirconia based composite tweezers with flexural strength of greater than 700 MPa and resistivities of 106-109 Ohm-cm. Further, the material showed a residual magnetic flux of up to 14 Gauss.
U.S. Pat. No. 6,136,232 (2000) reports that some perovskite type oxides can be mixed with stabilized zirconia to prepare electrostatic dissipative ceramics. The patent also describes that other perovskite type oxides react with zirconia to form zirconates, thus not satisfactory for electrostatic dissipative ceramics. The patent though reports only volume resistivities as a key property.
U.S. Pat. No. 6,274,524 (2001) claims a semiconductive zirconia body formed under oxidative conditions, comprising 60 to 90 weight % of ZrO2 including stabilizing agent, said zirconia body having no more than 2% by weight Al2O3, containing greater than 10 weight % to 40 weight % of one kind or more of oxides of Fe, Co, Ni, and Cr as conductivity giving agents, having a three-point flexural strength of at least 580 MPa, and having a volume specific resistance of 106 to 109 Ohm-cm.
The above prior art attempts at control of ESD have not been fully successful in satisfying either current ESD dissipative requirements or the predicted future requirements for ESD dissipative ceramics, for the following reasons:
1. The Prior Art Lacks Sufficient High Density
ESD dissipative ceramic compositions of the prior art to date have not had sufficiently high density (i.e., 97-99% T.D.). This is due to processing techniques used in the prior art, which have not included hot isostatic pressing (hereinafter xe2x80x9cHIPingxe2x80x9d). The prior art has typically not made hot isostatically pressed (hereinafter xe2x80x9cHIPedxe2x80x9d) materials due to thermochemical instabilities which during a typical HIPing process. The conductivity modifiers employed in the prior art ESD dissipative ceramics often undergo a thermochemical reduction during a HIPing at high temperature (1200xc2x0-1500xc2x0 C.) and high argon pressure (100-250 MPa) in a HIP using graphite heating elements. The thermochemical reaction during the HIPing reduces transitional metallic oxides to oxygen deficient metallic oxides by releasing gases (CO2 or O2) from the body to the surface resulting in bloating and cracking of ceramic bodies.
Pressureless sintered ESD dissipative ceramics have a number of pores (pits) on the surfaces and inside of bodies. For instance, sintered ceramics with 99% TD may have a number of pores of 0.1-10 xcexcm diameters. Assuming a square sintered plate with dimensions of 1 cmxc3x971 cmxc3x970.25 cm, 99% T.D., and monosized pores of 1 xcexcm diameter the number of exposed pores on the surface are 10 millions. Often these pores trap particulate debris and contaminants inside, thus difficult to clean and act as the source of contamination during the use of such components in clean room environments. Further, the cleaning of ceramics with residual pores is difficult compared to that of substantially pore free ceramics by HIPing. On the other hand, the number of pores on HIPed ceramic plate with the same dimension and 99.9% T.D. would be 1 million reducing the number of defect (residual pores) to {fraction (1/10)} of the sintered plates.
In addition, the residual pores may interact with diamond grits of grinding wheels during the machining into final shapes and dimensions resulting in damaged spots, which is a potential source of particulate debris over the life time of such components.
It is, therefore, highly desirable to prepare a ceramic composition which is easy to densify by a conventional hot isostatic press (HIP) to achieve fully dense, structurally reliable ceramic components.
2. The Prior Art Lacks Sufficient High Strength
The HIPing also further eliminates strength limiting cavities and pores in ceramic bodies resulting in structurally stronger and more reliable components. It is well known that the flexural strength of ceramic components can be improved after HIPing by greater than 20% up to over 50%. Accordingly, HIPing is a preferred processing step to prepare high strength, mechanically reliable components and enable a thinner cross section and complicated shapes.
In summary, HIPable ESD dissipative ceramics are advantageous for an improved contamination control in the fab due to the absence of residual pores.
3. The Prior Art Lacks Choices of Color
Modern manufacturing techniques involve a number of automated processing steps. The efficiency of vision system in such an automated manufacturing system is critical. The high productivity of vision system can be achieved by a fast optical recognition from differences in colors and contrast. The color of magnetic recording heads is substantially black due to the color of slider material (AlTiC, a composite of titanium carbide and alumina). Therefore, a xe2x80x9cnon-blackxe2x80x9d ESD dissipative material is highly preferred in order to ensure higher productivity in vision systems. The color of other microelectronic components may vary. Therefor it is desirable to have various color capabilities with ESD dissipative ceramic tools and fixtures. Most prior art ESD dissipative ceramics commercialized to date have been either black or substantially dark.
4. The Prior Art Lacks Tunability
The proper ESD dissipation for a specific application can be obtained by a material with a specific volume/surface resistivity. However, there exist a number of different applications requiring a broad range of surface resistivities as reported. In the prior art, the control of volume and surface resistivities has been obtained mainly by a compositional variation in ESD dissipation. Therefore, one should prepare various compositions of materials to satisfy various application needs, each of which may require different resistivities. In addition, the amount of resistivity modifier used in forming composites often tends to make appreciable changes in other material properties. The design engineer thus tends to modify the resistivity specification for the application based on test results. Accordingly, there is a need for a material having a tunable resistivity.
5. The Prior Art Lacks Sufficient Nonmagnetic Properties
For some applications, especially in GMR head manufacturing, it is desirable to have the lowest magnetic susceptibility of the ESD dissipative material. For other applications to measure electromagnetic performance of microelectronic devices requires no interference from fixtures. Most transition metal based resistivity modifiers exhibit a substantial magnetic susceptibility. It is reported in PCT Publication No. WO 98/49121 that a residual magnetic flux density of up to 16 Gauss is satisfactory for general ESD dissipative applications. However, there is a need for a substantially nonmagnetic material to be used in tools for electromagnetic measurements. Accordingly, for some applications, it is preferable to have an ESD dissipative material made with one or more substantially nonmagnetic resistivity modifiers.
The surface resistivity is not the only measure to determine the performance of ESD dissipative materials. In general, it is very desirable to have a fast dissipation of static charges from the contact at the surface. A more precise measure is a charge decay time in ms as described in the literature.
PCT Publication No. WO 98/49121 reports that an acceptable decay time from 1000 to 100 V is 0.1-20 seconds. Further, they reported that materials outside of that range are not suitable for ESD dissipative ceramics. However, it has been noted that a decay time of less than 0.1 second is preferred for most ESD dissipative applications.10